Web Logs Finally Meet Sim City
l0rd writes "A good piece on wired says :
A few games of Roller Coaster Tycoon don't usually translate into productive work, but for one developer the diversion planted the seed for making website analysis more intuitive.
Several years after playing those inspirational games, Robert Savage came up with VisitorVille, a website-traffic analysis package that essentially crosses the DNA of SimCity with that of the traditional chart- and graph-centric tools businesses have long been using.
Screenshots included."
Nice try, but since it's web only it's not going to handle email :-(.
:-(.
Looks like it works by putting a tracking image on your web site that resides on their servers and then using that to track remotely. Clever since it means you don't have to install any software on the web site or have control over your web server. On the other hand it would be a bit of a pain to edit all those pages. I'll have to dig deeper to see if it works with web sites that are all dynamic.
I have to say that I like the idea enough that I may well exhume my Windows machine to give it a go. Pity there's no Mac or Linux version
D
I like this idea. This could be a great replacement for the feeling you get when a physical customer walks into your store.
Is it feasible to just run VisitorVille on a PC or a big screen in your virtual store's office/room? I would enjoy watching a visitor walk around my city, go through various buildings all while I'm writing up product descriptions and working on site design. This could really give you a sense of how your business is growing, as well.
Has anyone actually used this product, yet?
Or on the corporte lan where user Joe has a 'house' and all of a sudden cars and people are jamming around it (he just emailed a link to his beta web project stored on his local PC).
And the BOFH could stomp through as King Kong and wreak havoc on Jane's mail-merge (since she attached a 5MB file instead of linking to it).
If not already posted, check this summary here: visual summary
Ok, so who's going to use perl/php with Ming modules to do this? (or something better of course).
I used to work for Webtrends doing technical support (yecch!) so I feel like I have some good insight into web analysis and it's pitfalls and benefits, as well as the types of people who use it. The biggest problems by far that were encountered was setup-configuration and graph interpretation. I think that both of these problems will be increased in a program like this, and that the cool factor provided by the model will not offset these problems for most serious webmasters.
I see a couple of problems with setup and configuration, but the biggest without doubt is "can it handle dynamic pages?" Is it able to discern the difference by pages when that difference is controlled by a URI query? What if the dynamic parameters are passed in with a POST? Will this require the tracking on each page to be modified? Many large companies use dynamic websites, so this could be a serious barrier.
As far as interpretation problems go, I think it's pretty cool that this software is able to give graphic metaphors for traffic on a web site, but it's hard to use abstract metaphors when doing business or web traffic analysis.
I think that this is going to be a tool, almost exclusively, of small websites that are able to tweak their web pages on a whim (unlike large companies are able to do, in most cases), which makes the price point even more of a problem. Thirty bucks a month?! That's a lot of money for someone who's running a small site, it could be more than their hosting fees.
It's a cool idea, and I like to see the virtual world evolving, but I don't think that this is going to do well.
Why should I argue rationally with someone being irrational? I'll just mock them instead.